View metadata, citation and similar papers at core.ac.uk brought to you by CORE
provided by Elsevier - Publisher Connector
Biochimie 93 (2011) 1701e1709
Contents lists available at ScienceDirect
Biochimie
journal homepage: www.elsevier.com/locate/biochi
Research paper Substrate specificity of kallikrein-related peptidase 13 activated by salts or glycosaminoglycans and a search for natural substrate candidates
Douglas Andrade a, Diego M. Assis a, Jorge A.N. Santos a, Fabiana M. Alves a, Izaura Y. Hirata a, Mariana S. Araujo b, Sachiko I. Blaber c, Michael Blaber c, Maria A. Juliano a, Luiz Juliano a,*
a Department of Biophysics, Escola Paulista de Medicina, Universidade Federal de São Paulo, Rua Três de Maio 100, 04044-020 São Paulo, Brazil b Department of Biochemistry, Escola Paulista de Medicina, Universidade Federal de São Paulo, Rua Três de Maio 100, 04044-20 São Paulo, Brazil c Department of Biomedical Sciences, College of Medicine, Florida State University, Tallahassee, FL 32306-4300, USA
article info abstract
Article history: KLK13 is a kallikrein-related peptidase preferentially expressed in tonsils, esophagus, testis, salivary Received 28 April 2011 glands and cervix. We report the activation of KLK13 by kosmotropic salts and glycosaminoglycans and Accepted 31 May 2011 its substrate specificity by employing a series of five substrates derived from the fluorescence resonance Available online 12 June 2011 energy transfer (FRET) peptide Abz-KLRSSKQ-EDDnp. KLK13 hydrolyzed all these peptides only at basic 0 residues with highest efficiency for R; furthermore, the S3 to S2 subsites accepted most of the natural Keywords: amino acids with preference also for basic residues. Using a support-bound FRET peptide library eight Kinin peptide substrates were identified containing sequences of proteins found in testis and one with myelin Protease Histatins basic protein sequence, each of which was well hydrolyzed by KLK13. Histatins are salivary peptides Kosmotropic salts present in higher primates with broad antifungal and mucosal healing activities that are generated from Myelin basic protein the hydrolysis from large precursor peptides. KLK13 efficiently hydrolyzed synthetic histatin 3 exclu- sively at R25 (DSHAKRHHGYKRKFHEKHHSHRGYR25YSNYLYDN) that is the first cleavage observed inside the salivary gland. In conclusion, the observed hydrolytic activities of KLK13 and its co-localization with its activators, glycosaminoglycans in the salivary gland and high concentration of sodium citrate in male reproductive tissues, indicates that KLK13 may play a role in the defense of the upper digestive apparatus and in male reproductive organs. Ó 2011 Elsevier Masson SAS. Open access under the Elsevier OA license.
1. Introduction preferences in glandular epithelia, tonsils, esophagus, trachea, testis, salivary glands, vagina and cervix [8e10]. The physiological The tissue kallikrein members represent a cluster of 15 serine roles of KLK13 remain unknown, in contrast to KLK1 that is peptidases of chymotrypsin-like S1A family clan PA(S) localized in involved in kinin release [11]; KLK3 hydrolyzes seminogelin I and II the human chromosome locus 19q13.4, and present a high degree in semen and contributes to its liquefaction after ejaculation of identity [1e7]. Human kallikrein-related peptidase 13 (KLK13) is [12,13]; KLK4 in prostate activates KLK3 [14,15], is modulated by one of these peptidases that is expressed in many tissues with a specific zinc binding site [16] and is involved in dental enamel formation [17]; KLK6 activates PAR2 and cleaves glutamatergic receptors and myelin basic protein [18e22]; KLK5 and KLK7 are Abbreviations: Q-EDDP, (glutamyl-N-[ethylenediamine] 2,4-dinitrophenyl); Abz, involved in skin desquamation (for a recent review see [23]) and ortho-aminobenzoic acid; HOBt, N hydroxybenzotriazole; TBTU, 2-(1H-benzo- KLK8 participates in human epidermis and sweat in a proteolytic fl triazole-1-yl)-1,1,3,3-tetramethylaminium tetra uorborate; NMM, N-methyl-mor- cascade contributing to the skin barrier [24]. The associations of pholine; (PMC), 2,2,5,7,8-pentamethylchroman-6-sulfonyl; Trt, trityl; DIPEA, diisopropylethylamine; TFA, Trifluor acetic acid; EDT, ethanodithiol; DMF, dime- KLK13 with some pathologies were earlier reported, and include thylformamide; NMM, 4-methylmorpholine; DCM, dichloromethane; GAGs, psoriasis vulgaris and atopic dermatitis [25] (where the peptidase is glycosaminoglycans; FRET, fluorescence resonance energy transfer; LMWK, low expressed in significant amounts), it is observed to be a favorable molecular weight kininogen; MCA, methyl coumarin amide; MALDI-TOF, Matrix- prognosis marker in ovarian cancer [26], is expressed in high levels assisted laser desorption/ionization-Time of flight; LCMS, liquid chromatography/ in some types of salivary gland tumors [27], is down regulated mass spectrometry. * Corresponding author. Tel.: þ55 11 5576 4450; fax: þ55 11 5575 9617. in testicular cancer [28,29] and is also reported to degrade the E-mail address: [email protected] (L. Juliano). extracellular matrix [30].
0300-9084 Ó 2011 Elsevier Masson SAS. Open access under the Elsevier OA license. doi:10.1016/j.biochi.2011.05.037 1702 D. Andrade et al. / Biochimie 93 (2011) 1701e1709
Detailed analyses of substrate specificity of peptidases are essen- We further tried to find other possible candidates as natural tial to identify their roles in mammalian organisms, to aid in the substrates for KLK13 using a support-bound FRET peptide library development of selective assays, and to guide the design of novel prepared on PEGA resin (a beaded polyethylene glycol dimethyla- inhibitors. The substrate specificity of KLK13 was earlier investigated crylamide copolymer) by the process of split-combine synthesis, by screening of fluorogenic 7-amino-4-carbamoylcoumarin positional which results in a single peptide sequence on each resin bead scanning-synthetic combinatorial libraries modified at the P1 to P4 [43e46]. The FRET peptide bound to a resin bead when hydrolyzed positions [31] andKLK13wasobservedtohaveselectivityoftheS1 by the peptidase turned the bead fluorescent; such beads were subsite for R, preference for hydrophobic amino acids by the S2 and S4 manually picked up and the remained peptide in the bead was subsites, and S3 selectivity mainly for R (Schechter and Berger sequenced by Edman degradation methodology. The identified nomenclature [32]). Though these synthetic combinatorial libraries peptides were synthesized and the cleavage sites by KLK13 were have provided valuable information about the specificity of numerous subsequently determined in solution and the kinetic parameters for peptidases [33,34] they are restricted to the non-prime side of the their hydrolysis were evaluated. peptidases, its application can be impaired for peptidases where the active site requires extended prime site substrates. In addition, 2. Materials and methods cleavages of these substrates other than at the amino site of the 7- amino-4-methylcoumarin group (peptidyl-Y-AMC)arenotdetected 2.1. Recombinant KLK13 with this library. Glycosaminoglycans (GAGs) and kosmotropic salts such as Mature KLK13 was expressed and purified from a baculovirus/ sodium citrate and sodium sulfate were reported to strongly acti- insect cell line system as previously described [22].Briefly, the KLK13 vate the hydrolytic activities of KLK3 and KLK6 [18,35,36] in gene which encodes for the mature form of KLK13 was inserted in contrast to KLK1 that is inhibited by salts [37]. In the present work frame, C-terminal of an enterokinase propeptide. To facilitate the we examined the modulation of salts and GAGs on hydrolytic purification procedure, a 6 His-tag was also included on the N- 0 activity of KLK13 and its S3 to S2 subsite substrate specificity taking terminal side of the enterokinase propeptide. The purification took as reference the fluorescence resonance energy transfer (FRET) advantage of the His-tag by Ni-NTA affinity chromatography. After peptide Abz-KLRSSKQ-EDDnp (Abz ¼ ortho-aminobenzoic acid and purification, KLK13 was activated by extended (i.e. 24 h) incubation EDDnp ¼ N-[2,4-dinitrophenyl]-ethylenediamine), which was with enterokinase at 1:100 ratio of enterokinase:pro-KLK13 (w/w). designed based on previously reported subsite requirements of Hydrolysis of the enterokinase propeptide proceeded to apparent KLK1 and KLK6 [18]. Five series of analogs of this peptide were completion and the released propeptide and enterokinase were synthesized with substitution of each amino acid, except Gln, and removed from the mature KLK13 by size exclusion chromatography. assayed as substrates of KLK13. The final yield of purified mature KLK13 was typically 20e25 mg per Histatins are a class of salivary peptides present in higher liter of culture. The final purity of mature KLK13 was assessed by primates, rich in basic amino acids [38] and have broad antifungal Coomassie Brilliant Blue staining of a 15% SDS-PAGE gel, N-terminal activity [39e41]. The systematic search by tandem mass spectros- sequencing, and mass spectrometry, and yielded a purity of at least copy of human saliva showed that the generation of histatin 3- 98% with no visible degradation (Fig. 1A) and the correct mature N- related peptides resulted mainly from the action of trypsin-like terminus. The purified, mature form of KLK13 remained a single chain proteases, and the first cleavage being at R25 (DSHAKRHHGYKRKF- and migrated as closely-spaced doublet bands under both reducing HEKHHSHRGYR25YSNYLYDN) [42]. The selectivity of KLK13 for and non-reducing conditions. The results of N-terminal sequencing hydrolysis at R and K and its co-localization with histatins in salivary revealed that both bands had identical termini consistent with glands [10] prompted us to examined the hydrolytic activity of mature KLK13 (VLNTXGTSGFLP). The 5th cycle of the reaction did not KLK13 on synthetic histatin 3 that is mainly processed in human give an amino acid assignment, and this is presumably due to saliva into histatin 6 (DSHAKRHHGYKRKFHEKHHSHRGYR25) and glycosylation of Asn. The doublet bands were separated by Phe- 25 then to histatin 5 by removing of C-terminal R [42]. Due to nomenex C4 reverse phase HPLC (Fig. 1-B) and analyzed by MALDI- selectivity of KLK13 to cleave at basic amino acids, the kininogenase TOF mass spectrometry (Fig. 1-C), and the peak of each fragment activity was also evaluated using low molecular weight (LMW) presented molecular masses 29,213 and 30,270 Da, respectively. kininogen, and a synthetic human kininogen fragment, Abz-MIS- These values are higher than the calculated mass of un-modified LMKRPPGFSPFRSSRI-NH2 that contains a bradykinin sequence. KLK13 presumably due to the result of heterogeneous glycosylation
Fig. 1. (A) Purified recombinant KLK13 analysis by 15% SDS-PAGE in reducing conditions. Lanes: 1-Molecular weight marker; 2-Mature form of KLK13 with doublet fragments; 3- and 4-Mature forms of KLK13 isolated by C4. (B) C4 reverse phase HPLC separation of doublet fragment. (C) MALDI-TOF analysis of KLK13 fragments. D. Andrade et al. / Biochimie 93 (2011) 1701e1709 1703
(with the higher-mass form containing approximately 6 additional was at 38 C) with 0.5 mM heparin or 1.5 M sodium citrate. The hexose units). Both fragments exhibited similar enzymatic activity enzymes were pre-incubated in the assay buffer for 3 min before against Tos-GPR-AMC. Molar concentration of active KLK13 was the addition of substrate. Fluorescence changes were monitored determined by titration with MUGB (4-Methylumbelliferyl p-guani- continuously at lex ¼ 320 nm and lem ¼ 420 nm. When fluorogenic dinobenzoate hydrochloride) by spectrofluorimetric tritation as MCA peptides were used, the excitation and emission wavelengths previously described [47]. Bovine myelin basic protein was purchased were changed to lex ¼ 380 and lem ¼ 460 nm, respectively. The from (AbD Serotec, Raleigh, NC, USA). enzyme concentrations for initial rate determinations were chosen at a level intended to hydrolyze less than 5% of the added substrate 2.2. Peptide synthesis over the time course of data collection. The slope of the generated fluorescence signal was converted into micromoles of substrate All FRET peptides and histatin-3 were obtained by solid-phase hydrolyzed per minute based on a calibration curve obtained from peptide synthesis as previously described [48,49] and using the the complete hydrolysis of each peptide. Fmoc-procedure in an automated bench-top simultaneous multiple solid-phase peptide synthesizer (PSSM 8 system from Shimadzu,- 2.6. Support-bound FRET peptide library screening Tokyo, Japan). The peptides were synthesized in TGR-resin (loading 0.2 mmol/g) using HBTU/HOBt as coupling reagent and the cleavage For all assays, the library beads were washed with water (3 ) of peptide-resin was accomplished with TFA:anisol:EDT:water and the assay buffer (3 ) before the addition of the enzyme. The (85:5:3:7). All peptides obtained were purified by semi-preparative reactions were stopped by dilution with 3 M HCl, and the mixtures HPLC on an Econosil C-18 column. The molecular weight and purity were washed thoroughly until pH 5e6 was reached. The beads (94% or higher) of synthesized peptides were checked by amino acid were transferred to a glass dish and inspected by fluorescence analysis and MALDI-TOF mass spectrometry, using a Microflex e LT microscopy (Stereomicroscope Stemi-Zeiss), and the fluorescent mass spectrometer (Bruker e Daltonics, Billerica, MA, USA). Stock beads were collected and transferred to a TFA-treated cartridge solutions of the peptides were prepared in DMF and the concen- filter for on-resin sequence analysis. The amino acid sequence and trations were measured spectrophotometrically using the molar cleavage point were determined by Edman degradation using extinction coefficient of 17,300 M 1 cm 1 at 365 nm. a PPSQ/23 protein sequencer (Shimadzu, Japan). KLK13 was assayed as follows: 20 mg of resin in 50 mM Tris, 1 mM EDTA, pH 7.5, at 2.3. Synthesis of support-bound FRET peptide library 25 C for 5 h with 75 nM of enzyme. The identified peptide sequences susceptible to hydrolysis by KLK13 were synthesized as The syntheses of libraries were carried out manually as previ- FRET Abz-peptidyl-Q-EDDnp peptides and assayed in solution with ously described [46]. Briefly, the libraries were synthesized on 1 g of the enzyme. PEGA1900 resin [50] in a 20 column Teflon synthesis block, using protected Fmoc amino acids. The resin was evenly distributed in the 2.7. Digestion of myelin basic Protein by KLK13 20 wells of the Teflon synthesis block, and the Fmoc groups were removed. Prior to coupling the Fmoc amino acids (1 equiv.) were Bovine myelin basic protein (MBP) was added to KLK13 at pre-activated with HOBt (1 equiv.), TBTU (1 equiv.) and NMM a 2000:1 mass ratio in 50 mM Tris and 1 mM EDTA, pH 7.5. This (2 equiv.) in DMF (1 ml) for 6 min and the activated amino acids mixture was incubated at 35 C, and time points were taken at 2, 4, were added, one to each of the 20 wells. After the completion of the 8, and 12 h. The MBP and degradative fragments were resolved coupling, the block was filled with DMF up to 1 cm above the top of using Tricine SDS-PAGE (16.5%) and developed by Comassie-Blue the wells and inverted, and the resin was mixed vigorously by and two bands were isolated and the N-terminal sequenced by agitation for 30 min in the mixing chamber. The block was again Edman degradation to identify the cleavage sites. inverted, evenly distributing the resins into the wells for washing and removal of the Fmoc group. This procedure was repeated for 2.8. Kininogenase activity of KLK13 the incorporation of all the randomized positions. After the randomized positions, the Fmoc-K(Abz-Boc) and Fmoc-K(Dnp) Low molecular weight human kininogen (LK) was purchased were incorporated. The side chain protecting groups were from Calbiochem. The ability of KLK13 to cleave the human LK was removed by treatment with a mixture of TFA:thioanisole:ethane- evaluated incubating 1 ml of LK (100 mg/ml) and KLK13 (1.8 nM) in dithiol:water (87:5:5:3) for 8 h. The resin was washed with 95% reaction mixtures containing 50 mM Tris, 0e1.5 M sodium citrate, acetic acid (4 ), DMF (4 ), 5% DIPEA in DMF (3 ), DMF (3 ), DCM pH 7.5, at 35 C for 4 h. Ethanol (3:1, v/v) was added, and the (6 ) and dried under vacuum. mixture was centrifuged at 1000 g for 15 min. The kinin content in the supernatant was measured by radioimmunoassay, as previ- 2.4. Glycosaminoglycans ously described [52] KLK1 was also assayed as comparative control.
Size-defined (12,000 Da) bovine lung heparin (The Upjohn Co.) 3. Results was prepared by using a size exclusion column approach [51]; heparan sulfate (16,000 Da) from bovine lung was a generous gift 3.1. Conditions for high KLK13 activity from Dr. P. Bianchini (Opocrin Research Laboratories, Modena, Italy); dermatan sulfate (12,000 Da) and chondroitin sulfate The FRET peptide Abz-KLRSSKQ-EDDnp was taken as the initial (25,000 Da) were purchased from Seikagaku Kogyo Co. (Tokyo, substrate to search for the best conditions required for efficient KLK13 Japan). hydrolytic activity. The peptide Abz-KLRSSKQ-EDDnp was hydrolyzed only at the ReS bond, and this activity of KLK13 at pH 7.5 was highly 2.5. Kinetic measurements activated by sodium citrate or sodium sulfate, as shown in Fig. 2A. High activation was also observed with heparin up to 0.5 mM(Fig. 2-B) The FRET peptides were assayed in a Shimadzu RF-1501 spec- and chondroitin and dermatan sulfate also activated KLK13 but at trofluorometer, at 35 C. The assays were performed in 50 mM Tris, concentration range 20e40 mM (data not showed). KLK13 bound 1 mM EDTA, pH 7.5 and 35 C (the temperature of highest activity to heparineSepharose resin and was eluted with 100 mM NaCl, 1704 D. Andrade et al. / Biochimie 93 (2011) 1701e1709
Fig. 2. (A) Activation of KLK13 by sodium citrate (C) and sodium sulfate (,). (B) Activation of KLK13 by heparin (C) pH profiles of KLK13 activity in the presence of sodium citrate (:), heparin (-) and only with buffer (C). The reactions were performed in 50 mM Tris, 1 mM EDTA, pH 6.5e9.0, using as substrate 7.5 mM Abz-KLRSSKQ-EDDnp, and [E] ¼ 3.8 nmol at 35 C. indicating a significant favorable electrostatic interaction between has a particular preference for K in contrast to R, as indicated by the heparin and KLK13 that was not observed with KLK1 and KLK6. The kcat/Km value for the hydrolysis of the peptide Abz-KLRYKSKQ- pH profiles of KLK13 activities in the presence of 1 M sodium citrate EDDnp that is 170-times higher than that of the peptide Abz- and heparin are showed in Fig. 2-C. The KLK13 activity was reduced KLRYRSKQ-EDDnp. The second best substrate of this series contains 0 close to zero in the presence of 50 mM Tris HCl (with no sodium N and the peptides with F, S and H at the P1 position were still citrate or heparin). The presence of NaCl, KCl, MgCl2 and CaCl2 up to efficiently hydrolyzed. Not unexpectedly, the peptide with X ¼ Pat 0 1 M concentration did not activate the enzyme. the S1 subsite was completely resistant to hydrolysis due to the imide bond of P. The kinetic parameters for hydrolysis of the series 0 Abz-KLRSXKQ-EDDnp (Table 3) also show that the S2 subsite of 3.2. Substrate specificity of KLK13 using FRET peptides derived KLK13 does not have restricted specificity but does present pref- from Abz-KLRSSKQ-EDDnp erence for the basic amino acid R, and the resistance to hydrolysis of 0 the peptide Abz-KLRSEKQ-EDDnp indicates that the S2 of KLK13 The peptide Abz-KLRSSKQ-EDDnp was taken as reference, and does not accepted negative charge. a series of five related FRET peptides were synthesized and assayed as substrates for KLK13 in 50 mM Tris at pH 7.5 and 35 Cinthe presence of 0.5 mM heparin, and for some peptides the assays were 3.3. Screening of support-bound FRET peptide library done in the presence of 1.5 M sodium citrate. The specificity of the S1 subsite was explored with the peptide series Abz-KLXSSKQ-EDDnp The peptides selected from the screening of this library were (where X ¼ R, K, F, Y, M, L, V, W, D, E, G, H, I, N, P, Q, S, and T). Only the synthesized as FRET Abz-peptidyl-Q-EDDnp peptides and assayed peptides Abz-KLRYSSKQ-EDDnp and Abz-KLKYSSKQ-EDDnp (i.e. with KLK13. Their sequences and velocities of hydrolysis in the X ¼ RorX¼ K) were hydrolyzed and only at the peptide bonds ReS presence of 1 M sodium citrate are shown in Table 4. The suscep- and KeS, respectively. The kinetic parameters of these hydrolyses in tibility to hydrolysis by KLK13 of the selected peptides was higher the presence of heparin and sodium citrate at pH 7.5 are presented in the presence of sodium citrate than in the presence of 50 mM e in Table 1. All other peptides were resistant in both conditions up to Tris HCl, and most of these peptides were cleaved at the carboxyl side of R. Only two peptides were cleaved at K and each with lower 60 nM of KLK13. These results indicate that the S1 subsite of KLK13 has a very restricted specificity for basic amino acids. The resistance velocities. It is noteworthy that the basic amino acids have signif- to hydrolysis of peptide Abz-KLNSSKQ-EDDnp contrasts with the icant frequency in all positions besides the P1 position and eight out of twenty sequences were found in proteins from testis. reported hydrolysis of peptides containing N at the P1 position in the screening of fluorogenic 7-amino-4-carbamoylcoumarin positional One human myelin basic protein sequence (GPVKKRNM) was scanning-synthetic combinatorial libraries [31]. found as a potential substrate of KLK13 in the support-bound fi The FRET peptides in the series Abz-XLRSSKQ-EDDnp, Abz- peptide library that was con rmed by the hydrolysis of the FRET KXRSSKQ-EDDnp, Abz-KLRXSKQ-EDDnp and Abz-KLRSXKQ-EDDnp, were synthesized in order to explore the specificity of the subsites 0 0 Table 1 S3,S2,S1 and S2 , respectively. The kinetic parameters for their Kinetic parameters for the hydrolysis by KLK13 of the reference peptides Abz- hydrolysis by KLK13 are shown in Tables 2 and 3. All the peptides of KLRSSKQ-EDDnp and Abz-KLKSSKQ-EDDnp in the presence of heparin and sodium the four series were hydrolyzed only at the ReS (or ReX) bond. citrate. 1 1 The parameters for the hydrolysis of the series Abz-KXRSSKQ- Activator kcat (s) Km (mM) kcat/Km (mM s) EDDnp (Table 2) shows that the S2 subsite of KLK13 did not Abz-KLRYSSKQ-EDDnp present restricted specificity; however, higher kcat/Km values were 0.5 mM heparin 3.2 0.1 2.2 0.1 1455 80 observed with substrates containing basic or hydrophobic amino 1.5 M sodium citrate 3.4 0.2 1.2 0.05 2834 204 acids at the P2 position. Similarly, an S3 subsite preference for basic Y and hydrophobic residues was observed as demonstrate from the Abz-KLK SSKQ-EDDnp 0.5 mM heparin 0.50 0.03 31 216 3 parameters of hydrolysis of the series Abz-XLRSSKQ-EDDnp 1.5 M sodium citrate 0.10 0.01 5.3 0.3 19 2 (Table 2). However, the peptide with the negatively charged The reactions were performed in 50 mM Tris, 1 mM EDTA, pH 7.5, 0.5 mM heparin residue E at the P position was resistant to hydrolysis. 3 and [KLK13] ¼ 3e10 nmol at 35 C. The parameters and the standard errors were The kinetic parameters for the hydrolysis of the peptide series obtained from three determinations. The errors associated to k /K were obtained 0 cat m Abz-KLRXSKQ-EDDnp (Table 3) show that the S1 subsite of KLK13 by error propagation of the kcat and Km errors. D. Andrade et al. / Biochimie 93 (2011) 1701e1709 1705
Table 2 Table 3 Kinetic parameters for the hydrolysis by KLK13 of the series of peptides Abz- Kinetic parameters for the hydrolysis by KLK13 of the series of peptides Abz- KXRSSKQ-EDDnp and Abz-XLRSSKQ-EDDnp in the presence of heparin (i.e. evalu- KLRXSKQ-EDDnp and Abz-KLRSXKQ-EDDnp in the presence of heparin (i.e. evalu- 0 0 ating the specificity of S2 and S3 subsites, respectively). ating the specificity of S1 and S2 subsites, respectively).